What problem does it solve? Teams deploying applications on Render often misconfigure PostgreSQL connections, hit connection limits, or lose data by deleting databases without exports. This Skill provides authoritative guidance on connection strings, storage autoscaling, HA, read replicas, and backup safety for Render Managed Postgres. ## Core Features & Use Cases - Connection Management: Explains internal vs external URLs, TLS requirements, Blueprint fromDatabase wiring, and application-side pooling since Render provides no built-in pooler. - Capacity & Availability Planning: Covers storage autoscaling rules (~90% trigger, 12-hour cooldown, no shrink), RAM-based connection limits, HA prerequisites, and up to 5 read replicas with declarative Blueprint sync warnings. - Backup & Recovery: Details daily snapshots, PITR, pg_dump/pg_restore workflows, and the critical rule that backups are not retained after database deletion. - Use Case: A developer hits "too many connections" errors in production. The Skill identifies the RAM-based connection limit, recommends application-side pooling or PgBouncer, and suggests read replicas for read-heavy traffic. ## Quick Start Ask the AI to explain how to connect your Render web service to a Render Postgres database using the internal connection string and Blueprint fromDatabase wiring.